Kahneman interview - putting a premium on error detection

Xavier Prida dr.xavier.prida at GMAIL.COM
Sat Feb 9 20:52:10 UTC 2019


I must comment on Mike Bruno's insightful taxonomy of error classification.
I had developed and we had utilized a very similar classification in a
primary project of STEMI system of care for ECG interpretation and
assignment for emergent cath lab activation. Our taxonomy was:

1.) Interpretive Errors - "I didn't see that"

2.) Cognitive Errors - "I saw that, but, I didn't know that(or what it
meant)"

3.) System Errors - "Something happened that should not have, or, something
did not happen, but should have."

Xavier

* praesent superare odio  (r**ise above)*

Xavier E. Prida MD FACC FSCAI
Assistant Professor of Medicine
Program Director Cardiology Fellowship Training
USF Morsani College of Medicine
Department of Cardiovascular Sciences
2 Tampa General Circle
STC 5 th Floor
Tampa, Fl 33606
813 259 0992(O)
813 831 0721(H)
813 245 3143(C)


On Sat, Feb 9, 2019 at 12:16 PM Gerrit Jager <gerrit.jager at planet.nl> wrote:

> I fully agree with Mike, but I like to add another category; the no-fault
> errors. The sensitivity of a diagnostic test is almost never 100%, e.g. a
> chest X-ray has too little spatial resolution to detect small nodules and
> in most patients missing them is an unavoidable misdiagnosis related to the
> limitations of the test. But abandon chest X-ray and perform CT in every
> patient is not the alternative.
> (It is a different story when the selection of an imaging test is
> inappropriate.)
>
> So the message is that we should always be aware and prepared that we may
> be wrong and be willing to reconsider a diagnosis.
>
> This is not easy as excellently described by Kathryn Schultz in her book
> “Being Wrong, adventures in the margin of error”.  She explains why we are
> so bad at imaging that we are mistaken.
>
> Gerrit Jager
>
> Op 07-02-19 15:59, Bruno, Michael <mbruno at PENNSTATEHEALTH.PSU.EDU>
> schreef:
>
> Thanks, Nelson,
>
> Yes, I agree—process changes, but also cognitive changes and an
> accommodation to our basic human biology.  I like to think of errors as
> falling into three categories: (1) those which are due to faults in our
> work processes, (2) t hose which are due to faults in our thinking, and (3)
> those which occur simply because of how we are made.  Each of these will be
> amenable to different solutions—but only error detection after the fact
> addresses all three.
>
> While there are a few diagnoses which, if missed, lead to immediate and
> irreparable harm, most do not.  If we learn to be inclined to expect,
> search for, and rapidly detect errors we will prevent harm MOST of the
> time.  In our experience in radiology, we do on regular occasion detect
> errors which have been present for several hours, such as when a finding is
> missed overnight but picked up on re-review in the morning.  In virtually
> all of these, patient harm was averted by correcting the diagnosis a few
> hours later.
>
> All the best,
>
> Mike
>
>
>
> *From:* Nelson Toussaint [mailto:ntoussaint at tamarac.com
> <ntoussaint at tamarac.com>]
> *Sent:* Thursday, February 07, 2019 8:36 AM
> *To:* 'Society to Improve Diagnosis in Medicine'; Bruno, Michael
> *Subject:* Kahneman interview - putting a premium on error detection
>
>
>
>
> February 7, 2019
> 8:16 AM
>
>
> *Michael *
> In Aerospace, it is anticipated that this behavior will occur.  Equipment
> will fail and people will make mistakes!  So, a concept of *Accommodation*
> is built into the process to counteract those faults that could lead to
> serious harm.  In some cases you can detect the errors and in some the
> consequences are in play before you can confirm a detection.  So where real
> harm can appear, the Accommodation needs to be somewhat active before/as
> the fault occurs.
>
> An example is the Ground Proximity Warning System, which estimates if the
> airplane may soon fly into terrain (navigation fault of the pilot or
> equipment).  This is only a warning and the pilot must take an action to
> change course.
>
> Early detection of diagnostic error is good; but it still may not be in
> time to limit the potential harm.  A more robust process would include
> methods where the "conclusions" are reviewed before a diagnosis is
> complete.  One that comes to mind is what many of us use quite often -
> "what does this seem like to you"; a discussoin amongst involved parties
> such as the attending physician - radiologist/pathologist – patient.  This
> Accommodation can serve to cause the “experts” to pause and reasses their
> findings.  It does not give a foolproof solution, but still allows the
> experts to continue to form the diagnosis.
>
> This is just a thought, but it seems to me the process changes are the
> best way to attack this problem in order to get broad participation.  The
> trick will be how to modify the process without penalizing the many
> straight-forward cases.
>
>  Nelson Toussaint
>
> TAMARAC LLC
> 860-844-0199
> ntoussaint at tamarac.com <mailto:ntoussaint at tamarac.com
> <ntoussaint at tamarac.com>>
>
>
> *From:* Bruno, Michael [mailto:mbruno at PENNSTATEHEALTH.PSU.EDU
> <mbruno at PENNSTATEHEALTH.PSU.EDU>]
> *Sent:* Tuesday, February 05, 2019 12:41 PM
> *To:* IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG
> *Subject:* Re: [IMPROVEDX] Kahneman interview - putting a premium on
> error detection
>
> Yes, thanks, Art!
>
> It seems to me that while some errors are preventable, others may well be
> inevitable. The perceptual errors we were talking about at that session at
> DEM are probably biologically rooted (*i.e.,* secondary to neurocognitive
> brain network functions) and are thus essentially outside of our conscious
> control. They are an example of an entire class of errors that flow from
> “how we are made,” our biology and our evolution, and so they will not
> respond to the usual interventions, such as cognitive de-biasing, adult
> learning/CME, or even mindfulness.
>
> So it seems to me that, for these types of errors at least, there needs to
> be a premium placed on *early error detection*, so that errors can be
> more promptly detected and corrected before any patient harm is done.  That
> is the value of double-reading in radiology, and it may be an avenue where
> AI turns out to be particularly helpful in the years ahead.
>
> All the best,
>
>
> *Michael A. Bruno, M.D., M.S., F.A.C.R.*
> Professor of Radiology & Medicine
> Vice Chair for Quality & Patient Safety
> Chief, Division of Emergency Radiology
> Penn State Milton S. Hershey Medical Center
> ( (717) 531-8703  |  6 (717) 531-5737
> *** mbruno at pennstatehealth.psu.edu <mailto:mbruno at pennstatehealth.psu.edu
> <mbruno at pennstatehealth.psu.edu>>  |
>
>
> *****E-Mail Confidentiality Notice*****
> This message (including any attachments) contains information intended for
> a specific individual(s) and purpose that may be privileged, confidential
> or otherwise protected from disclosure pursuant to applicable law.  Any
> inappropriate use, distribution or copying of the message is strictly
> prohibited and may subject you to criminal or civil penalty.  If you have
> received this transmission in error, please reply to the sender indicating
> this error and delete the transmission from your system immediately.
>
>
>
> *From:* Art Papier [mailto:apapier at VISUALDX.COM <apapier at VISUALDX.COM>]
> *Sent:* Sunday, February 03, 2019 2:34 PM
> *To:* IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG
> *Subject:* Re: [IMPROVEDX] Kahneman interview
>
> Thanks for sharing!   Great interview with many fascinating threads, and
> some new thoughts on the randomness of error, and how all error is not due
> to cognitive bias.  Towards the conclusion we hear that Dr. Kahneman in not
> a believer in cognitive debiasing…. saying essentially we are too busy
> making errors to recognize that we are making errors.  He asserts we should
> be thinking about how we recognize other peoples errors.  In made me think
> of the session at DEM on perceptual errors in diagnostic imaging and the
> very positive role of second reads in radiology….whether by humans or AI to
> recognize errors.  Perhaps either co-decision making, or AI “second
> opinions” is an area we should all be further exploring.
> Best
> Art
>
> Art Papier MD
> CEO VisualDx
> Associate Professor of Dermatology and Medical Informatics
> University of Rochester College of Medicine
>
>
> ------------------------------
>
>
>
> *From:* Xavier Prida <dr.xavier.prida at GMAIL.COM>
> *Sent:* Sunday, February 03, 2019 8:59 AM
> *To:* IMPROVEDX at LIST.IMPROVEDIAGNOSIS.ORG
> *Subject:* [IMPROVEDX] Kahneman interview
>
>
> On "thinking again"- not changing your mind, error, and bias - the latter
> two are not always linked.
>
>
>
>
> https://onbeing.org/programs/daniel-kahneman-why-we-contradict-ourselves-and-confound-each-other-jan2019/
> <
> https://urldefense.proofpoint.com/v2/url?u=https-3A__onbeing.org_programs_daniel-2Dkahneman-2Dwhy-2Dwe-2Dcontradict-2Dourselves-2Dand-2Dconfound-2Deach-2Dother-2Djan2019_&d=DwMGaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=XZJky8Jx0OuETXcWpBMhx9j_wSYpSZPDVXdInJ5O9gQ&m=UrbAzwfML-iEIuCykgw2Fqn20pE6WrN96IDUMbiOHMs&s=3X_IjCf1rsQokLp5q2wzaZqF1pXrJky5MKUV2tVppnQ&e=>
>
>
>
>
> XEP
>
>
>
>
> *praesent superare odio (rise above) *
>
>
> Xavier E. Prida MD FACC FSCAI
>
> Assistant Professor of Medicine
>
> Program Director Cardiology Fellowship Training
>
> USF Morsani College of Medicine
>
> Department of Cardiovascular Sciences
>
> 2 Tampa General Circle
>
> STC 5 th Floor
>
> Tampa, Fl 33606
>
> 813 259 0992(O)
>
>
>
>
> ------------------------------
>
>
>
>
> To unsubscribe from IMPROVEDX: click the following link:
>
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1
> <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__list.improvediagnosis.org_scripts_wa-2DIMPDIAG.exe-3FSUBED1-3DIMPROVEDX-26A-3D1&d=DwMGaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=XZJky8Jx0OuETXcWpBMhx9j_wSYpSZPDVXdInJ5O9gQ&m=UrbAzwfML-iEIuCykgw2Fqn20pE6WrN96IDUMbiOHMs&s=PBCG5hjllVPSANZQoKcMqMD2ZiVXDpIMwesAc65piDE&e=>
>
> or send email to: IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG
>
> Visit the searchable archives or adjust your subscription at:
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?INDEX <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__list.improvediagnosis.org_scripts_wa-2DIMPDIAG.exe-3FINDEX&d=DwMGaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=XZJky8Jx0OuETXcWpBMhx9j_wSYpSZPDVXdInJ5O9gQ&m=UrbAzwfML-iEIuCykgw2Fqn20pE6WrN96IDUMbiOHMs&s=OeKMCzd2zeZiCtm1TGcYnGbJmDibC0OzIARvu7ewz1o&e=>
>
>
> Moderator:David Meyers, Board Member, Society for Improving Diagnosis in
> Medicine
>
> To learn more about SIDM visit:
> http://www.improvediagnosis.org/ <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.improvediagnosis.org_&d=DwMGaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=XZJky8Jx0OuETXcWpBMhx9j_wSYpSZPDVXdInJ5O9gQ&m=UrbAzwfML-iEIuCykgw2Fqn20pE6WrN96IDUMbiOHMs&s=p9DtK4wdXsBYva9iE7z3eA3n1OqmC6On22QUAHNrHbo&e=>
>
>
>
> ------------------------------
>
>
>
>
> To unsubscribe from IMPROVEDX: click the following link:
>
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1
> <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__list.improvediagnosis.org_scripts_wa-2DIMPDIAG.exe-3FSUBED1-3DIMPROVEDX-26A-3D1&d=DwMFaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=n5GC0R3w9jcHuCKxwuM7Hf8RiNko4O90e45rH34twIE&m=iRZnYLBXW_98lAxXmISASl94HhCeEaWW7QJXaxDu_tw&s=G_nkbwM3zjXrbSFsZspyy4Q7UwS14TMKbehSGipHfEE&e=>
>
> or send email to: IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG
>
> Visit the searchable archives or adjust your subscription at:
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?INDEX <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__list.improvediagnosis.org_scripts_wa-2DIMPDIAG.exe-3FINDEX&d=DwMFaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=n5GC0R3w9jcHuCKxwuM7Hf8RiNko4O90e45rH34twIE&m=iRZnYLBXW_98lAxXmISASl94HhCeEaWW7QJXaxDu_tw&s=PTKjDDT0_CcAdZdHkueSf9c3hGE5bCtfYg0RsL5U-2A&e=>
>
>
> Moderator:David Meyers, Board Member, Society for Improving Diagnosis in
> Medicine
>
> To learn more about SIDM visit:
> http://www.improvediagnosis.org/ <
> https://urldefense.proofpoint.com/v2/url?u=http-3A__www.improvediagnosis.org_&d=DwMFaQ&c=_FmMnDvUH5queZcSmOuBzHZMbp7E7EwtGwv5cxxnTj0&r=n5GC0R3w9jcHuCKxwuM7Hf8RiNko4O90e45rH34twIE&m=iRZnYLBXW_98lAxXmISASl94HhCeEaWW7QJXaxDu_tw&s=yA9UySZFBC0mFBmFXfuW14G9sh--1HsK31u2ZoL-EAw&e=>
>
>
> ------------------------------
>
>
> To unsubscribe from IMPROVEDX: click the following link:
>
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1
>  or send email to: IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG
>
>
> Moderator:David Meyers, Board Member, Society for Improving Diagnosis in
> Medicine
>
> To learn more about SIDM visit:
> http://www.improvediagnosis.org/
>
>
> ------------------------------
>
>
> To unsubscribe from IMPROVEDX: click the following link:
>
> http://list.improvediagnosis.org/scripts/wa-IMPDIAG.exe?SUBED1=IMPROVEDX&A=1
> or send email to: IMPROVEDX-SIGNOFF-REQUEST at LIST.IMPROVEDIAGNOSIS.ORG
>
> Moderator:David Meyers, Board Member, Society for Improving Diagnosis in
> Medicine
>
> To learn more about SIDM visit:
> http://www.improvediagnosis.org/






Moderator: David Meyers, Board Member, Society to Improve Diagnosis in Medicine


HTML Version:
URL: <../attachments/20190209/3c07b560/attachment.html> ATTACHMENT:
Name: image.png Type: image/png Size: 2281 bytes Desc: not available URL: <../attachments/20190209/3c07b560/attachment.png> ATTACHMENT:
Name: image.png Type: image/png Size: 6043 bytes Desc: not available URL: <../attachments/20190209/3c07b560/attachment-0001.png>


More information about the Test mailing list